CN111582152A - Method and system for identifying complex event in image - Google Patents
Method and system for identifying complex event in image Download PDFInfo
- Publication number
- CN111582152A CN111582152A CN202010376777.3A CN202010376777A CN111582152A CN 111582152 A CN111582152 A CN 111582152A CN 202010376777 A CN202010376777 A CN 202010376777A CN 111582152 A CN111582152 A CN 111582152A
- Authority
- CN
- China
- Prior art keywords
- events
- image
- event
- complex
- identifying
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 34
- 230000006870 function Effects 0.000 claims abstract description 29
- 238000010586 diagram Methods 0.000 claims description 24
- 238000004590 computer program Methods 0.000 claims description 9
- 238000005516 engineering process Methods 0.000 abstract description 6
- NVNSXBXKNMWKEJ-UHFFFAOYSA-N 5-[[5-(2-nitrophenyl)furan-2-yl]methylidene]-1,3-diphenyl-2-sulfanylidene-1,3-diazinane-4,6-dione Chemical compound [O-][N+](=O)C1=CC=CC=C1C(O1)=CC=C1C=C1C(=O)N(C=2C=CC=CC=2)C(=S)N(C=2C=CC=CC=2)C1=O NVNSXBXKNMWKEJ-UHFFFAOYSA-N 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 3
- 238000012544 monitoring process Methods 0.000 description 3
- 230000005540 biological transmission Effects 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 125000002015 acyclic group Chemical group 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 239000002131 composite material Substances 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/20—Scenes; Scene-specific elements in augmented reality scenes
Abstract
The invention provides a method for identifying complex events in an image, which relates to the field of image identification and comprises the following steps: s1, identifying the objects contained in the image and extracting the identification probability value of each object; s2, connecting the objects through Boolean logic to form a plurality of Boolean functions for judging whether the events expressed in the images occur; s3, calculating the occurrence probability values of the events corresponding to the different Boolean functions in the step S2 one by one according to the recognition probability values, judging whether the events occur or not, and forming recognition results whether the events occur or not and whether the events can be recognized or not; s4, taking the event as a root node, taking the identification result of the event as an output edge to construct a decision graph, and judging the complex event containing a plurality of events so as to identify whether the complex event in the image occurs or can be identified; the invention solves the problem of difficult complex operation during the identification of the complex events in the image, and has stronger adaptability on the image identification technology.
Description
Technical Field
The invention relates to the technical field of image recognition, in particular to a method and a system for recognizing complex events in an image.
Background
With the rapid development of emerging information technologies such as internet of things, cloud computing, big data, artificial intelligence and the like, the monitoring of physical environment through intelligent electronic products becomes possible. By arranging a series of sensors or monitoring instruments in a physical environment, the state of the physical environment can be known from the sensed data or image video data, such environment being an intelligent environment in which events reflect changes in the state of the observed object. Sensors and monitoring instruments generate a large amount of data at all times, and the surrounding environment is reflected in an intelligent manner. The data are accurately counted, analyzed and synthesized, the implicit information in the data is extracted, reasoning is carried out on the information level, and the events occurring in the environment are identified, so that decision can be better made and the reaction can be quickly taken.
The processing of the information usually adopts artificial intelligence analysis, and the artificial intelligence technology can be used for classifying and mining massive data such as sensing data, videos, pictures, sounds and the like, and the accuracy and efficiency are very high, but in the process of processing the video image data, the video image data is usually damaged, the data is polluted, the data contains complex scenes of complex types, or the data which is not generated in some complex environments and the like, and the application under the complex conditions usually needs a large amount of high-quality data sets to accumulate and learn and train and optimize a learning model for a long time so as to be suitable for the application under the complex environments or in complex events, and the identification accuracy is ensured. This limits the convenience and versatility of the machine learning algorithm in practical application scenarios, and the computational complexity and operational difficulty of the machine learning algorithm in such complex scenarios will increase by a factor of two.
The binary decision diagram is a directed acyclic diagram, comprises root nodes and child nodes, the nodes are connected by 1 or 0 branches, the child nodes respectively use 1 and 0 to represent two non-intersecting logic states, the binary decision diagram is a data structure used for expressing a Boolean function, is a decision analysis method for evaluating project risks and judging feasibility of the project risks, can be widely used in many fields, comprises digital chip design and system reliability analysis, and can be used for analyzing data and predicting. Particularly, when the binary decision diagram is used for solving the fault tree in the field of system reliability, the binary decision diagram technology shows the advantages of low operation complexity and high efficiency, but the decision diagram is mainly applied to the fields of reliability and engineering control, and because no logical operation relation exists among image data, the decision diagram has no specific application in the field of image recognition.
Disclosure of Invention
Aiming at the defects in the prior art, the invention provides a method for identifying complex events in an image, which solves the problems of complex operation and high operation difficulty in identifying the events in the image in the prior art.
According to an embodiment of the present invention, a method for identifying a complex event in an image comprises the following steps:
s1, identifying the object: identifying objects contained in the image and extracting identification probability values of the objects;
s2, establishing an event occurrence judgment logic: the objects are connected through Boolean logic to form a plurality of Boolean functions for judging whether the events expressed in the images occur or not;
s3, event identification: calculating occurrence probability values of the events corresponding to the different Boolean functions in the step S2 one by one according to the recognition probability values, so as to judge whether the events occur and form recognition results whether the events occur or not and whether the events can be recognized or not;
s4, complex event recognition: and taking the event as a root node, taking the identification result of the event as an output edge to construct a decision graph, and judging the complex event containing a plurality of events so as to identify whether the complex event in the image occurs or can be identified.
Further, the events contain states or interrelationships between objects or combinations of objects to reflect dynamic or procedural descriptions. Further, the decision graph comprises a binary decision graph and/or a ternary decision graph.
Further, the decision graph comprises a binary decision graph and/or a ternary decision graph.
Furthermore, the operation relationship between the binary decision diagram nodes adopts binary and/or logic operation.
Furthermore, the recognition result of the ternary decision diagram to the image is divided into three categories: and the identification result is true, false and unrecognizable, the multiple ternary decision graphs are combined into a more complex ternary decision graph through binary logic operation, and the operation among the nodes of the ternary decision graph follows the operation expression among the nodes of the binary decision graph.
Furthermore, before the boolean function is established, an and/or logical relationship is given to each object in the image, so as to form the boolean function for judging whether an event occurs.
According to an embodiment of the present invention, there is also provided an electronic device including a memory having a computer program stored therein and a processor configured to perform any one of the steps of the above-mentioned method for identifying a complex event in an image when running.
According to an embodiment of the present invention, there is also provided a computer-readable storage medium having stored therein a computer program arranged to, when executed, perform any one of the steps of the above-described method of identifying a complex event in an image.
According to an embodiment of the present invention, there is also provided a system for identifying a complex event in an image, the system including the electronic device or the computer-readable storage medium.
The technical principle of the invention is as follows: the method comprises the steps of extracting a characteristic value of each object in image data by utilizing logic operation of a decision graph and a deep learning technology, wherein the characteristic value represents a probability value of each object, establishing a corresponding Boolean function for each event to enable a complex event in the image data to be recognized by a machine, finally constructing the decision graph with an image recognition function according to the extracted characteristic value and the established Boolean function, and finally judging the probability of the complex event in the image data by node operation of the decision graph.
Compared with the prior art, the invention has the following beneficial effects: the method has the advantages that the logical relation is given to each object in the image data, the logical relation is represented by the decision diagram, the occurrence probability of the complex event in the image data is finally recognized, the operation complexity of the image recognition technology in the complex event is simplified by using the logical operation of the decision diagram, the recognition operation rate is improved, the operation is simple, and in addition, a large number of data sets do not need to be prepared in advance for learning training, so the method has stronger adaptability in various complex and changeable scenes.
Drawings
Fig. 1 is a flowchart illustrating a method for identifying a complex event in an image according to an embodiment of the present invention.
Fig. 2 is a representation of event recognition for the UCF101 data set according to this embodiment.
FIG. 3 is a schematic diagram of 2 ternary decision graphs merged into a new ternary decision graph according to an AND or binary logic operation according to another embodiment of the present invention.
Detailed Description
The technical solution of the present invention is further explained with reference to the drawings and the embodiments.
Example 1:
the embodiment of the invention provides a method for identifying a complex event in an image, and as shown in fig. 1, the method for identifying a complex event in an image according to the embodiment of the invention is a flow chart, and comprises the following steps:
step S1, performing multi-object recognition on the standard data set 'UCF 101' by using a deep learning VGGNet-19 depth model, and extracting recognition probability values of different objects, wherein the events comprise states or interrelations among objects or object combinations and are used for reflecting dynamic or process description things;
the complex event as in the present embodiment is an event such as "roller skating performance", "wedding", "feeding animal", "carpenter art", "fishing", "birthday party" contained in each image data in the data set "UCF 101", and the object identified by the image is each object in the image.
In this embodiment, the recognition probability values of some of the objects extracted from the data set "UCF 101" are respectively: human (X)10.984), wedding dress (X)20.876), candle (X)30.765), flower (X)40.897), balloon (X)50.654), lawn (X)60.786), vehicle (X)70.976), camera (X)8＝0.689)。
Step S2, using heuristic method to link each object through Boolean logic to form several Boolean functions to judge whether the event expressed in the image occurs;
since the boolean expression for failure judgment in the reliability field is generally obtained by a long-term experience summary of technicians, in this embodiment, the selected boolean expression is first selected according to actual conditions and through a knowledge base and the routine experience of the technicians, an object associated with the event is selected, the selected object is assigned with a boolean logic relationship, then a large number of complex event image data positive samples are selected, all possibly occurring boolean functions are listed under the existing boolean logic condition relationship, the occurrence probabilities of events corresponding to different boolean functions are calculated one by one, and the boolean expression with the highest corresponding probability value is selected as the final boolean function.
In order to select a Boolean function suitable for a wedding event, the embodiment combines all objects in the image with the sumThe logical relations are combined to form a Boolean function, all combinations of the Boolean functions which can be formed are calculated, all the possible Boolean functions are calculated by considering the basic objects, the objects which possibly appear and the interference factors of the occurrence of the wedding event and utilizing the identification probability values of the objects obtained in the step S1, and then the Boolean expression with the highest corresponding probability value is selected as the final Boolean function F of the wedding event (X is equal to F is the final Boolean function of the wedding event)1+(X1*X2*X8)+X4+X5+X6)+(X1*X2*X7*X8)。
Step S3, calculating the occurrence probability values of the events corresponding to the different Boolean functions in the step S2 one by one according to the identification probability values, so as to judge whether the events occur and form an identification result whether the events occur or not and whether the events can be identified or not;
in this embodiment, a binary decision diagram is first established by using a consistent equation according to the identification characteristic values and the boolean functions in the above steps, and the binary decision diagram in this embodiment has two output branches, which are two mutually exclusive states, i.e., an unrecognizable and recognizable state, i.e., a "0" edge and a "1" edge in the binary decision diagram, and a root node represents an event, and a child node is represented by 0 or 1.
The operations between the nodes of the decision graph established in this embodiment all follow binary logic operations F1 ═ F, F0 ═ 0, F +1 ═ 1, and F +0 ═ F, where "+" represents an or relationship and "+" represents an and relationship.
Step S4, taking the event as a root node, taking the identification result of the event as an output edge to construct a decision graph, and judging the complex event containing a plurality of events so as to identify whether the complex event in the image occurs or not and whether the complex event can be identified or not;
based on the binary decision diagram for image recognition established in step S3, a ternary decision diagram with an image recognition function is constructed, and the identification result of the ternary decision diagram constructed in this embodiment is unrecognizable, true, and false, that is, the "0" side, the "1" side, and the "2" side in the ternary decision diagram. The root node X of the ternary decision diagram represents an event, three branches are three mutually exclusive states of the event respectively, and child nodes are represented by 0 or 1.
In the embodiment, the ternary decision diagram logical structure expression uses a quadruple CASE to represent F ═ x0·Fx＝0+x1·Fx＝1+x2·Fx＝2＝0·Fx＝0+1·Fx＝1+0·F x＝20+1+0 ═ CASE (X,0,1,0), where X0、x1、x2Respectively representing the states corresponding to the '0', '1', '2' edges of the root node X, F0、F1、F2Respectively represent the child nodes corresponding to the root node X in the states of "0" edge, "1" edge, and "2" edge, where X is in this embodiment0＝0、x1＝1、x2＝0。
As shown in fig. 2, the occurrence probability values of the events corresponding to the respective edges are calculated by the operation between the nodes of the ternary decision graph, and the user classifies the events in each image of the data set "UCF 101" by event type according to the event occurrence probability, and for example, the present embodiment obtains a "wedding" event boolean function in step S2 to reconstruct the decision graph, and then calculates the recognition results of the "wedding" event as "unrecognizable 0.870", "recognized true 0.100", and "recognized false 0.030" by the operation between the nodes of the decision graph.
Before step S2 is performed, the present embodiment forms a boolean function for determining whether an event has occurred by assigning an and logical relationship to each object in an image to express the relationship between the objects in the image.
Example 2:
according to needs, this implementation may be combined with embodiment 1, and in a CASE where a more complex event needs to be identified, according to a heuristic method, a plurality of ternary decision graphs may be combined into one more complex ternary decision graph by combining a recursive operation with a binary logic operation, in this embodiment, two events G and H to be identified, and a CASE expression G ═ CASE (X, G) of the ternary decision graph of the two events0,G1,G2)、H＝CASE(Y,H0,H1,H2)，
Where "◇" represents any of the "and or" bivariate logical operations, "W" represents the resultant event, X, Y represents the root nodes of events G and H, respectively, H0、H1、H2、G0、G1、G2The sub-nodes respectively representing events G and H, index (X), index (Y) represent X, Y node index values in the composite ternary decision diagram, and the present embodiment adopts the document [1 ] for the sorting mode of node index values]Sasao T.Ternary decisiondiagrams.Survey[C]The method mentioned in International Symposium on Multiple-valued logic. IEEE,2002. As shown in FIG. 3, the embodiment is described in index (X)<index (Y) the two events G and H are combined by AND/or binary logic.
It should be noted that, the above steps can be implemented by software or hardware, and for the latter, the following modes can be implemented, but are not limited to the following modes: the steps of the method are all positioned in the same processor; alternatively, the method steps may be located in different processors, in any combination.
Embodiments of the present invention also provide a computer-readable storage medium having a computer program stored thereon, wherein the computer program is arranged to perform the steps of any of the above-mentioned method embodiments when executed. In this embodiment, the computer-readable storage medium may include, but is not limited to: various media capable of storing computer programs, such as a usb disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a removable hard disk, a magnetic disk, or an optical disk.
Embodiments of the present invention also provide an electronic device comprising a memory having a computer program stored therein and a processor arranged to run the computer program to perform the steps of any of the above method embodiments. The electronic device may further include a transmission device and an input/output device, wherein the transmission device is connected to the processor, and the input/output device is connected to the processor.
It will be apparent to those skilled in the art that the steps of the present invention described above may be implemented by a general purpose computing device, they may be centralized on a single computing device, or distributed over a network of multiple computing devices, or they may be implemented by program code executable by a computing device, such that they may be stored in a storage device and executed by a computing device, and in some cases, the steps described in the embodiments may be executed out of order from that shown, or separately fabricated into individual integrated circuit modules, or fabricated into a single integrated circuit module, and the present invention is not limited to any specific combination of hardware and software.
Through the above description of the embodiments, those skilled in the art can clearly understand that the method according to the above embodiments can be implemented by software plus necessary general hardware platform. Based on such understanding, the technical solutions of the present invention may be embodied in the form of a software product, which is stored in a storage medium, such as ROM, RAM, magnetic disk, optical disk, and includes instructions for enabling a terminal device, such as a mobile phone, a computer, a server, or a network device, to execute the method according to the embodiments of the present invention.
In the embodiment, because the program operation of the invention during the event identification in the image only relates to binary logic operation, the complex logic relationship between objects in the image is simplified, the whole operation process is simpler, and the identification rate is improved.
Finally, the above embodiments are only for illustrating the technical solutions of the present invention and not for limiting, although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions may be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all of them should be covered in the claims of the present invention.
Claims (9)
1. A method for identifying complex events in an image, comprising the steps of:
s1, identifying the object: identifying objects contained in the image and extracting identification probability values of the objects;
s2, establishing an event occurrence judgment logic: the objects are connected through Boolean logic to form a plurality of Boolean functions for judging whether the events expressed in the images occur or not;
s3, event identification: calculating occurrence probability values of the events corresponding to the different Boolean functions in the step S2 one by one according to the recognition probability values, so as to judge whether the events occur and form recognition results whether the events occur or not and whether the events can be recognized or not;
s4, complex event recognition: and taking the event as a root node, taking the identification result of the event as an output edge to construct a decision graph, and judging the complex event containing a plurality of events so as to identify whether the complex event in the image occurs or can be identified.
2. A method of identifying complex events in images as claimed in claim 1, wherein the events contain states or interrelationships between objects or combinations of objects to reflect dynamic or procedural descriptive matter.
3. The method for identifying complex events in images as claimed in claim 1, wherein the decision graph comprises a binary decision graph and/or a ternary decision graph.
4. The method for identifying complex events in images as claimed in claim 3, wherein the operation relationship between the binary decision diagram nodes adopts binary and/or logic operation.
5. The method for identifying complex events in images according to claim 3, wherein the identification result of the ternary decision diagram for events is divided into three categories: and the identification result is true, false and unrecognizable, a plurality of ternary decision graphs are combined into a new ternary decision graph through binary logic operation, and the operation among the nodes of the new ternary decision graph follows the operation expression among the nodes of the binary decision graph.
6. The method of claim 1, wherein the Boolean function is configured to assign an AND/or logical relationship to each object in the image before the Boolean function is established.
7. An electronic device, characterized in that the electronic device comprises a memory in which a computer program is stored and a processor arranged to perform the steps of a method of identifying complex events in an image as claimed in any of the preceding claims 1-6 when executed.
8. A computer-readable storage medium, in which a computer program is stored which, when being executed, is configured to carry out the steps of the method for identifying complex events in an image according to any one of the preceding claims 1 to 6.
9. A system for identifying complex events in an image, comprising the electronic device of claim 7 or the computer-readable storage medium of claim 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010376777.3A CN111582152A (en) | 2020-05-07 | 2020-05-07 | Method and system for identifying complex event in image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010376777.3A CN111582152A (en) | 2020-05-07 | 2020-05-07 | Method and system for identifying complex event in image |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111582152A true CN111582152A (en) | 2020-08-25 |
Family
ID=72118805
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010376777.3A Pending CN111582152A (en) | 2020-05-07 | 2020-05-07 | Method and system for identifying complex event in image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111582152A (en) |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050162515A1 (en) * | 2000-10-24 | 2005-07-28 | Objectvideo, Inc. | Video surveillance system |
| US20070092143A1 (en) * | 2005-10-21 | 2007-04-26 | The Boeing Company | System, method and computer program product for adaptive video processing |
| US20100128984A1 (en) * | 2008-11-25 | 2010-05-27 | Microsoft Corporation | Labeling Image Elements |
| CN103294876A (en) * | 2013-07-08 | 2013-09-11 | 重庆松藻煤电有限责任公司 | Automatic analysis method of graphical accident tree for evaluating coal mine safety |
| US20140149376A1 (en) * | 2011-06-23 | 2014-05-29 | Cyber Ai Entertainment Inc. | System for collecting interest graph by relevance search incorporating image recognition system |
| US20150279182A1 (en) * | 2014-04-01 | 2015-10-01 | Objectvideo, Inc. | Complex event recognition in a sensor network |
| CN106650076A (en) * | 2016-12-14 | 2017-05-10 | 武汉理工大学 | Ternary decision graph-based universal analysis method for fault-tolerant system |
| CN107908740A (en) * | 2017-11-15 | 2018-04-13 | 百度在线网络技术（北京）有限公司 | Information output method and device |
| US20190147354A1 (en) * | 2017-11-13 | 2019-05-16 | International Business Machines Corporation | Event identification through machine learning |
| US20190236415A1 (en) * | 2018-01-30 | 2019-08-01 | Beamr Imaging Ltd. | Method of classification of an object and system thereof |
| US20200372625A1 (en) * | 2018-12-19 | 2020-11-26 | Aquifi, Inc. | Systems and methods for joint learning of complex visual inspection tasks using computer vision |
-
2020
- 2020-05-07 CN CN202010376777.3A patent/CN111582152A/en active Pending
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050162515A1 (en) * | 2000-10-24 | 2005-07-28 | Objectvideo, Inc. | Video surveillance system |
| US20070092143A1 (en) * | 2005-10-21 | 2007-04-26 | The Boeing Company | System, method and computer program product for adaptive video processing |
| US20100128984A1 (en) * | 2008-11-25 | 2010-05-27 | Microsoft Corporation | Labeling Image Elements |
| US20140149376A1 (en) * | 2011-06-23 | 2014-05-29 | Cyber Ai Entertainment Inc. | System for collecting interest graph by relevance search incorporating image recognition system |
| CN103294876A (en) * | 2013-07-08 | 2013-09-11 | 重庆松藻煤电有限责任公司 | Automatic analysis method of graphical accident tree for evaluating coal mine safety |
| US20150279182A1 (en) * | 2014-04-01 | 2015-10-01 | Objectvideo, Inc. | Complex event recognition in a sensor network |
| CN106650076A (en) * | 2016-12-14 | 2017-05-10 | 武汉理工大学 | Ternary decision graph-based universal analysis method for fault-tolerant system |
| US20190147354A1 (en) * | 2017-11-13 | 2019-05-16 | International Business Machines Corporation | Event identification through machine learning |
| CN107908740A (en) * | 2017-11-15 | 2018-04-13 | 百度在线网络技术（北京）有限公司 | Information output method and device |
| US20190236415A1 (en) * | 2018-01-30 | 2019-08-01 | Beamr Imaging Ltd. | Method of classification of an object and system thereof |
| US20200372625A1 (en) * | 2018-12-19 | 2020-11-26 | Aquifi, Inc. | Systems and methods for joint learning of complex visual inspection tasks using computer vision |
Non-Patent Citations (3)
| Title |
|---|
| 万国春等: "《系统芯片（SOC）设计方法与实践》", 31 December 2016, 同济大学出版社 * |
| 胡涛等, 国防工业出版社 * |
| 邢留冬等: "《可靠性新技术丛书 动态系统可靠性理论》", 30 September 2019 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110147722A (en) | A kind of method for processing video frequency, video process apparatus and terminal device | |
| CN109816032B (en) | Unbiased mapping zero sample classification method and device based on generative countermeasure network | |
| CN112434721A (en) | Image classification method, system, storage medium and terminal based on small sample learning | |
| GB2604552A (en) | Fusing multimodal data using recurrent neural networks | |
| US10467547B1 (en) | Normalizing text attributes for machine learning models | |
| CN111931809A (en) | Data processing method and device, storage medium and electronic equipment | |
| CN115048370B (en) | Artificial intelligence processing method for big data cleaning and big data cleaning system | |
| CN111400452A (en) | Text information classification processing method, electronic device and computer readable storage medium | |
| CN113342489A (en) | Task processing method and device, electronic equipment and storage medium | |
| CN115718846A (en) | Big data mining method and system for intelligent interactive network | |
| CN109783626A (en) | Problem generation method, intelligent Answer System, medium and computer system | |
| CN115344805A (en) | Material auditing method, computing equipment and storage medium | |
| CN114462582A (en) | Data processing method, device and equipment based on convolutional neural network model | |
| CN112418256A (en) | Classification, model training and information searching method, system and equipment | |
| CN113434857A (en) | User behavior safety analysis method and system applying deep learning | |
| CN116756327B (en) | Threat information relation extraction method and device based on knowledge inference and electronic equipment | |
| CN113144624A (en) | Data processing method, device, equipment and storage medium | |
| CN111582152A (en) | Method and system for identifying complex event in image | |
| CN111353577A (en) | Optimization method and device of multi-task-based cascade combination model and terminal equipment | |
| CN111259975B (en) | Method and device for generating classifier and method and device for classifying text | |
| CN115062779A (en) | Event prediction method and device based on dynamic knowledge graph | |
| CN114492366A (en) | Binary file classification method, computing device and storage medium | |
| CN113886547A (en) | Client real-time conversation switching method and device based on artificial intelligence and electronic equipment | |
| CN115587297A (en) | Method, apparatus, device and medium for constructing image recognition model and image recognition | |
| CN111459990A (en) | Object processing method, system, computer readable storage medium and computer device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200825 |